Planning the Future Electricity Mix
: Designing the Use of Artificial Neural Networks to Investigate Energy Potentials of Renewable Generation Technologies, Electric Vehicles, Their Use Cases

  • Michael Allison

Student thesis: Doctoral Thesis


The world is currently in the midst of a fourth major energy transition which is intended to reduce dependency on fossil fuels. This transition is motivated by the desire to move towards a more sustainable energy paradigm which is less harmful to the environment, and which will also increase the energy security of countries. Increasing levels of renewable technologies such as photovoltaic (PV) systems into the fuel mix of the global electricity generation sector and the electrification of the transport sector are essential to support the move to a sustainable energy paradigm.
Whilst electrifying the transport sector and increasing the penetration levels of PV can support the move to a sustainable energy paradigm, they also pose a major challenge for electricity network operators and their aging and overworked systems. These challenges are heightened for operators in the global south where electricity demand is predicted to increase exponentially this century due to ambitious economic and social development programs. One of the major challenges facing operators is predicting how these changes will affect patterns and peaking characteristics of load profiles especially as the rate and scale of change is unknown.
This research presents a new scalable computational method which is proven to be capable of synthetically generating load profiles of electricity networks which will inevitably become significantly more complex in the near future. A systematic design approach that can be used to ensure that an optimal model can be found for any unique load forecasting scenario is also presented and forms the basis of investigation of select future energy use cases.
Many countries in the global south are currently engaged in programmes that aim to exploit high indigenous renewable energy potential to meet forecasted increasing demand for electricity. A case study of Yangon City, Myanmar was used to investigate the suitability of using PV in these endeavours and to examine the diurnal variation in PV output and the effects of this variable output on local load demand profiles over the course of a year. The results of the study demonstrated a strong correlation between PV output and local load demand, meaning that there would be little grid support needed from non-renewable generation and storage technologies to accommodate increasing PV levels.
The output from PV systems at times need to be curtailed to prevent network conditions such as voltage rise. This curtailment negatively affects the financial viability of PV systems. A case study of three countries at different stages of economic development was carried out to investigate the efficacy of different low-cost smart grid solutions in reducing or even preventing PV curtailment. Results showed that updating grid codes alone can prevent curtailment in some locations. They also showed that combining different smart grid solutions for locations in the global south could reduce curtailment at all PV penetration levels.
Date of Award1 Nov 2022
Original languageEnglish
Awarding Institution
  • Teesside University

Cite this